We investigated the cardiorespiratory health effects of smoke exposure from the 1997 Southeast Asian Forest Fires among persons who were hospitalized in the region of Kuching, Malaysia. We selected admissions to seven hospitals in the Kuching region from a database of all hospital admissions in the state of Sarawak during January 1, 1995 and December 31, 1998. For several cardiorespiratory disease classifications we used Holt-Winters time-series analyses to determine whether the total number of monthly hospitalizations during the forest fire period (August 1 to October 31, 1997), or post-fire period (November 1, 1997 to December 31, 1997) exceeded forecasted estimates established from a historical baseline period of January 1, 1995 to July 31, 1997. We also identified age-specific cohorts of persons whose members were admitted for specific cardiorespiratory problems during January 1 to July 31 of each year (1995--1997). We compared Kaplan-Meier survival curves of time to first readmission for the 1997 cohorts (exposed to the forest fire smoke) with the survival curves for the 1995 and 1996 cohorts (not exposed, pre-fire cohorts). The time-series analyses indicated that statistically significant fire-related increases were observed in respiratory hospitalizations, specifically those for chronic obstructive pulmonary disease (COPD) and asthma. The survival analyses indicated that persons over age 65 years with previous hospital admissions for any cause (chi2(1df) = 5.98, p = 0.015), any cardiorespiratory disease (chi2(1df) = 5.3, p = 0.02), any respiratory disease (chi2(1df) = 7.8, p = 0.005), or COPD (chi2(1df) = 3.9, p = 0.047), were significantly more likely to be rehospitalized during the follow-up period in 1997 than during the follow-up periods in the pre-fire years of 1995 or 1996. The survival functions of the exposed cohorts resumed similar trajectories to unexposed cohorts during the post-fire period of November 1, 1997 to December 31, 1998. Communities exposed to forest fire smoke during the Southeast Asian forest fires of 1997 experienced short-term increases in cardiorespiratory hospitalizations. When an air quality emergency is anticipated, persons over age 65 with histories of respiratory hospitalizations should be preidentified from existing hospitalization records and given priority access to interventions.

Forest fires burned near the Hoopa Valley Indian Reservation in northern California from late August until early November in 1999. The fires generated particulate matter reaching hazardous levels. We assessed the relationship between patients seeking care for six health conditions and PM(10) exposure levels during the 1999 fires and during the corresponding period in 1998 when there were no fires. Multivariate logistic regression analysis indicated that daily PM(10) levels in 1999 were significant predictors for patients seeking care for asthma, coronary artery disease and headache after controlling for potential risk factors. Stratified multivariate logistic regression models indicated that daily PM(10) levels in 1999 were significant predictors for patients seeking care for circulatory illness among residents of nearby communities and new patients, and for respiratory illness among residents of Hoopa and those of nearby communities.

This study examines health effects resulting from landslides in Chuuk during Tropical Storm Chata'an in July 2002, and suggests strategies to prevent future mortality. In August 2002, we conducted a cross-sectional survey to identify risk factors for mortality during landslides, which included 52 survivors and 40 surrogates for 43 decedents to identify risk factors for death. Findings suggest that 1) females had a higher mortality rate from this event than males, and 2) children aged 5-14 years had a 10-fold increase in mortality when compared with annual mortality rates from all causes. Awareness of landslides occurring elsewhere and knowledge of natural warning signs were significantly associated with lower risks of death; being outside during landslides was not associated with reduced mortality. In Chuuk, improving communication systems during tropical storms and increasing knowledge of natural warnings can reduce the risk for mortality during landslides.

The question of how to test if collected data for a case-control study are misclassified was investigated. A mixed approach was employed to calculate the Bayes factor to assess the validity of the null hypothesis of no-misclassification. A real-world data set on the association between lung cancer and smoking status was used as an example to illustrate the proposed method.